An evaluation of patient-reported outcome measures in lower limb reconstruction surgery

Purpose To assess the measurement properties (acceptability, validity, reliability and responsiveness), of the MOS 36-Item Short-Form Health Survey (SF-36), the EQ-5D, the Short-Form McGill Pain Questionnaire (SF-MPQ) and the Musculoskeletal Functional Assessment Instrument (MFA), in patients who have undergone limb reconstruction surgery (LRS). Methods Four instruments measuring patient-reported outcome were completed at baseline and 12 months from surgery. Results 101 LRS patients were recruited with 95 responding at baseline and 71 at a 12-month follow-up. Response rates at baseline were over 94%. In three instruments, there was evidence of floor or ceiling effect, the exception being the EQ-5D. Cronbach's alpha statistics of internal consistency reliability were acceptable at >= 0.80 for all dimensions of the MFA, the SF-MPQ PRI(S) and seven of the SF-36 dimensions. When comparing mean changes in scores between baseline and 12 months, the most responsive measure was the SF-36 with an average Standardised Response Mean of 0.48 for those who reported their health as better. Statistically significant differences were observed between the health change groups ('worse', 'better' and 'same') for four dimensions of the SF-36, the two summary scores and the SF-6D. Conclusions Variation and poor performance of some of the instruments resulted in a recommendation of using the SF-36 and the SF-6D for LRS patients.